Downloads & Free Reading Options - Results

Damages to the nervous system are one of the health challenges. Using electrical stimuli because of the nervous system's electrical nature has been introduced as a solution to differentiate the stem cells successfully in recent years. Electrical stimulation (ES) has been used in various cell culture methods to grow stem cells such as nerve stem cells (NSCs), nerve differentiation, migration, and repair. Electrical stimulation mechanisms direct axon and neurite growth and cause directional cell migration, while magnetic fields cause neurogenesis and help NSC differentiate into functional neurons/nerve cells. Conductive nanomaterials have been utilized as functional scaffolds to provide mechanical support and biophysical signals to direct the growth and differentiation of nerve cells and form complex neural tissue patterns. Electrical signals may improve stem cell neurogenesis through activating specific ion channels, such as SCN1α. This article can be used as a checklist for ES work in stem cell research to expand using stem cells in clinical applications. This study revealed that electrical conductive materials and applying electrical and magnetic signals to stem cells could be a promising option in treating diseases of the nervous system, such as spinal cord injuries. , Parkinson's, and other diseases related to the nervous system.

Damages to the nervous system are one of the health challenges. Using electrical stimuli because of the nervous system's electrical nature has been introduced as a solution to differentiate the stem cells successfully in recent years. Electrical stimulation (ES) has been used in various cell culture methods to grow stem cells such as nerve stem cells (NSCs), nerve differentiation, migration, and repair. Electrical stimulation mechanisms direct axon and neurite growth and cause directional cell migration, while magnetic fields cause neurogenesis and help NSC differentiate into functional neurons/nerve cells. Conductive nanomaterials have been utilized as functional scaffolds to provide mechanical support and biophysical signals to direct the growth and differentiation of nerve cells and form complex neural tissue patterns. Electrical signals may improve stem cell neurogenesis through activating specific ion channels, such as SCN1α. This article can be used as a checklist for ES work in stem cell research to expand using stem cells in clinical applications. This study revealed that electrical conductive materials and applying electrical and magnetic signals to stem cells could be a promising option in treating diseases of the nervous system, such as spinal cord injuries. , Parkinson's, and other diseases related to the nervous system.

This article is from International Journal of Molecular Sciences , volume 13 . Abstract Macroporous polyacrylamide-grafted-chitosan scaffolds for neural tissue engineering were fabricated with varied synthetic and viscosity profiles. A novel approach and mechanism was utilized for polyacrylamide grafting onto chitosan using potassium persulfate (KPS) mediated degradation of both polymers under a thermally controlled environment. Commercially available high molecular mass polyacrylamide was used instead of the acrylamide monomer for graft copolymerization. This grafting strategy yielded an enhanced grafting efficiency (GE = 92%), grafting ratio (GR = 263%), intrinsic viscosity (IV = 5.231 dL/g) and viscometric average molecular mass (MW = 1.63 × 106 Da) compared with known acrylamide that has a GE = 83%, GR = 178%, IV = 3.901 dL/g and MW = 1.22 × 106 Da. Image processing analysis of SEM images of the newly grafted neurodurable scaffold was undertaken based on the polymer-pore threshold. Attenuated Total Reflectance-FTIR spectral analyses in conjugation with DSC were used for the characterization and comparison of the newly grafted copolymers. Static Lattice Atomistic Simulations were employed to investigate and elucidate the copolymeric assembly and reaction mechanism by exploring the spatial disposition of chitosan and polyacrylamide with respect to the reactional profile of potassium persulfate. Interestingly, potassium persulfate, a peroxide, was found to play a dual role initially degrading the polymers—“polymer slicing”—thereby initiating the formation of free radicals and subsequently leading to synthesis of the high molecular mass polyacrylamide-grafted-chitosan (PAAm-g-CHT)—“polymer complexation”. Furthermore, the applicability of the uniquely grafted scaffold for neural tissue engineering was evaluated via PC12 neuronal cell seeding. The novel PAAm-g-CHT exhibited superior neurocompatibility in terms of cell infiltration owing to the anisotropic porous architecture, high molecular mass mediated robustness, superior hydrophilicity as well as surface charge due to the acrylic chains. Additionally, these results suggested that the porous PAAm-g-CHT scaffold may act as a potential neural cell carrier.

Emotional Cognitive Neural Algorithms with Engineering Applications

Machine Learning + Neural Network Live Sessions Lectures By 5 Minutes Engineering

Engineering optimization problems involve minimizing some function subject to constraints. In areas such as aircraft optimization, the constraint equations may be from numerous disciplines such as transfer of information between these disciplines and the optimization algorithm. They are also suited to problems which may require numerous re-optimizations such as in multi-objective function optimization or to problems where the design space contains numerous local minima, thus requiring repeated optimizations from different initial designs. Their use has been limited, however, by the fact that development of response surfaces randomly selected or preselected points in the design space. Thus, they have been thought to be inefficient compared to algorithms to the optimum solution. A development has taken place in the last several years which may effect the desirability of using response surfaces. It may be possible that artificial neural nets are more efficient in developing response surfaces than polynomial approximations which have been used in the past. This development is the concern of the work.

We aim to provide readers with a state-of-the-art quantitative analysis of hydrogel-based nanocomposites used in neural tissue engineering. Readers will gain a holistic understanding of the nanostructure types used across various studies and obtain a comprehensive view of the methodologies employed for integration and outcome evaluation of nanostructure incorporation. Furthermore, we will highlight details regarding differentiation induction and assessment, which is crucial for the reproducibility of the experiments.

This third annual report provides an overview of the research activities carried on in the Center for Neural Engineering (CNE) comprising consortium partners: Tennessee State University (TSU), Meharry Medical College (MMC), Accurate Automation Corporation (AAC) and Oak Ridge National Laboratory (ORNL). A team of eight (8) researchers along with eight (8) undergraduates, six (6) graduates and one (1) Ph.D student from MMC conducted research at the Center. The Center conducted research in Neural computation using the experimental data on evoked potential from rat's auditor cortex, and on the performance of a neural network designed to perform a memory-based spatial navigation task, using data from hippocampal region. CNE also applied various neural network architectures and algorithms in the areas of auditory response, speech coding, neuromuscular signal decomposition and prothesis control, sensory motor control systems and intelligent aircraft control system. As a spin-off CNE was awarded a technology transfer research project from NASA in robotics under the STTR initiative.

Neural computing, as a paradigm of knowledge representation and information processing, has attracted tremendous enthusiasm and research interest recently. With advancing sophistication, neural computing technology has been successfully tailored for a wide range of applications, including some engineering fields. With the development of hardware based neural networks and neural computing theory, neural networks potentially provide efficient tools for solving some difficult engineering problems related to U.S. Army Construction Engineering Research Laboratories (USACERL) research. This report reviews and describes different types of neural networks including feedforward, feedback, and recurrent networks, their learning algorithms and recent developments. The emphasis is on the most frequently used multilayer feedforward neural networks. Representative publications on neural network applications to engineering problems related and/or of interest to research at USACERL, especially civil engineering problems, are also covered, and each modeling methodology identified. An extensive reference list is provided with each subject. The appendices list major technical journals dedicated to the theory and application of neural computing, some publicly available neural network simulators, and selected books and proceedings published on neural networks and genetic algorithms.

Talk given by Stéphane Deny of Facebook AI Research.  Given to the Redwood Center for Theoretical Neuroscience at UC Berkeley via Zoom.  There are two video files because due to technical problems, there was a break in the recording during the discussion after the main talk. Abstract: Why is visual information transmitted through many parallel channels in the optic nerve, with each channel encoding a different feature-map of the visual scene? Why do neurons in the retina prefer disk-shaped light dots and in the brain oriented lines? In this talk we will see how these simple questions can be investigated using artificial neural networks and a bit of maths. Bio: After completing my PhD in computational neuroscience at Pierre and Marie Curie University (Paris), and a postdoc in theoretical neuroscience in the department of Applied Physics of Stanford, I am currently working at the interface of neuroscience and artificial intelligence at Facebook AI Research in NYC. I develop mathematical theories inspired by physics and machine learning to understand the structure and organization of the brain, and also use insights from neuroscience to build more robust, flexible and energy-efficient AI systems.

Talk given by Stéphane Deny of Facebook AI Research.  Given to the Redwood Center for Theoretical Neuroscience at UC Berkeley via Zoom.  There are two video files because due to technical problems, there was a break in the recording during the discussion after the main talk. Abstract: Why is visual information transmitted through many parallel channels in the optic nerve, with each channel encoding a different feature-map of the visual scene? Why do neurons in the retina prefer disk-shaped light dots and in the brain oriented lines? In this talk we will see how these simple questions can be investigated using artificial neural networks and a bit of maths. Bio: After completing my PhD in computational neuroscience at Pierre and Marie Curie University (Paris), and a postdoc in theoretical neuroscience in the department of Applied Physics of Stanford, I am currently working at the interface of neuroscience and artificial intelligence at Facebook AI Research in NYC. I develop mathematical theories inspired by physics and machine learning to understand the structure and organization of the brain, and also use insights from neuroscience to build more robust, flexible and energy-efficient AI systems.

Talk given by Stéphane Deny of Facebook AI Research.  Given to the Redwood Center for Theoretical Neuroscience at UC Berkeley via Zoom.  There are two video files because due to technical problems, there was a break in the recording during the discussion after the main talk. Abstract: Why is visual information transmitted through many parallel channels in the optic nerve, with each channel encoding a different feature-map of the visual scene? Why do neurons in the retina prefer disk-shaped light dots and in the brain oriented lines? In this talk we will see how these simple questions can be investigated using artificial neural networks and a bit of maths. Bio: After completing my PhD in computational neuroscience at Pierre and Marie Curie University (Paris), and a postdoc in theoretical neuroscience in the department of Applied Physics of Stanford, I am currently working at the interface of neuroscience and artificial intelligence at Facebook AI Research in NYC. I develop mathematical theories inspired by physics and machine learning to understand the structure and organization of the brain, and also use insights from neuroscience to build more robust, flexible and energy-efficient AI systems.

Talk given by Stéphane Deny of Facebook AI Research.  Given to the Redwood Center for Theoretical Neuroscience at UC Berkeley via Zoom.  There are two video files because due to technical problems, there was a break in the recording during the discussion after the main talk. Abstract: Why is visual information transmitted through many parallel channels in the optic nerve, with each channel encoding a different feature-map of the visual scene? Why do neurons in the retina prefer disk-shaped light dots and in the brain oriented lines? In this talk we will see how these simple questions can be investigated using artificial neural networks and a bit of maths. Bio: After completing my PhD in computational neuroscience at Pierre and Marie Curie University (Paris), and a postdoc in theoretical neuroscience in the department of Applied Physics of Stanford, I am currently working at the interface of neuroscience and artificial intelligence at Facebook AI Research in NYC. I develop mathematical theories inspired by physics and machine learning to understand the structure and organization of the brain, and also use insights from neuroscience to build more robust, flexible and energy-efficient AI systems.

Engineering optimization problems involve minimizing some function subject to constraints. In areas such as aircraft optimization, the constraint equations may be from numerous disciplines such as transfer of information between these disciplines and the optimization algorithm. They are also suited to problems which may require numerous re-optimizations such as in multi-objective function optimization or to problems where the design space contains numerous local minima, thus requiring repeated optimizations from different initial designs. Their use has been limited, however, by the fact that development of response surfaces randomly selected or preselected points in the design space. Thus, they have been thought to be inefficient compared to algorithms to the optimum solution. A development has taken place in the last several years which may effect the desirability of using response surfaces. It may be possible that artificial neural nets are more efficient in developing response surfaces than polynomial approximations which have been used in the past. This development is the concern of the work.

The use of continuum models for the analysis of discrete built-up complex aerospace structures is an attractive idea especially at the conceptual and preliminary design stages. But the diversity of available continuum models and hard-to-use qualities of these models have prevented them from finding wide applications. In this regard, Artificial Neural Networks (ANN or NN) may have a great potential as these networks are universal approximators that can realize any continuous mapping, and can provide general mechanisms for building models from data whose input-output relationship can be highly nonlinear. The ultimate aim of the present work is to be able to build high fidelity continuum models for complex aerospace structures using the ANN. As a first step, the concepts and features of ANN are familiarized through the MATLAB NN Toolbox by simulating some representative mapping examples, including some problems in structural engineering. Then some further aspects and lessons learned about the NN training are discussed, including the performances of Feed-Forward and Radial Basis Function NN when dealing with noise-polluted data and the technique of cross-validation. Finally, as an example of using NN in continuum models, a lattice structure with repeating cells is represented by a continuum beam whose properties are provided by neural networks.

pages cm

pages cm

pages cm

This Fourth annual Report provides an overview of the research activities in the Center for Neural Engineering (CNE) comprising consortium partners: Tennessee State University (TSU), Caltech, Meharry Medical College (MMC), North East Ohio Universities College of Medicine (NEOUCOM), Oak Ridge National Laboratior (ORNL), and the University of Southwest Louisiana (USL). MMC and EOUCOM provided the experimental data. A team of eight (8) researchers along with (5) undergraduate students and twelve (12) graduate students conducted research at the Center. The CNE also supported a Ph.D student at NEOUCOM. CNE conducted research in various aspects of neural computation and applied these techniques to dynamic control aircraft, (helicopter), signal classification (image processing), spatial navigation (mobile robot) medical diagnosis and oscillatory hippocampal network (spatial information processing). The research led to the completion of three Masters Theses and three senior projects; one book chapter, four referenced chapters, six non-referenced conference papers and three new grants in the areas of communication, condition based maintenance and data acquisition system for medical diagnosis.

This Fourth annual Report provides an overview of the research activities in the Center for Neural Engineering (CNE) comprising consortium partners: Tennessee State University (TSU), Caltech, Meharry Medical College (MMC), North East Ohio Universities College of Medicine (NEOUCOM), Oak Ridge National Laboratior (ORNL), and the University of Southwest Louisiana (USL). MMC and EOUCOM provided the experimental data. A team of eight (8) researchers along with (5) undergraduate students and twelve (12) graduate students conducted research at the Center. The CNE also supported a Ph.D student at NEOUCOM. CNE conducted research in various aspects of neural computation and applied these techniques to dynamic control aircraft, (helicopter), signal classification (image processing), spatial navigation (mobile robot) medical diagnosis and oscillatory hippocampal network (spatial information processing). The research led to the completion of three Masters Theses and three senior projects; one book chapter, four referenced chapters, six non-referenced conference papers and three new grants in the areas of communication, condition based maintenance and data acquisition system for medical diagnosis.

This Fourth annual Report provides an overview of the research activities in the Center for Neural Engineering (CNE) comprising consortium partners: Tennessee State University (TSU), Caltech, Meharry Medical College (MMC), North East Ohio Universities College of Medicine (NEOUCOM), Oak Ridge National Laboratior (ORNL), and the University of Southwest Louisiana (USL). MMC and EOUCOM provided the experimental data. A team of eight (8) researchers along with (5) undergraduate students and twelve (12) graduate students conducted research at the Center. The CNE also supported a Ph.D student at NEOUCOM. CNE conducted research in various aspects of neural computation and applied these techniques to dynamic control aircraft, (helicopter), signal classification (image processing), spatial navigation (mobile robot) medical diagnosis and oscillatory hippocampal network (spatial information processing). The research led to the completion of three Masters Theses and three senior projects; one book chapter, four referenced chapters, six non-referenced conference papers and three new grants in the areas of communication, condition based maintenance and data acquisition system for medical diagnosis.

This Fourth annual Report provides an overview of the research activities in the Center for Neural Engineering (CNE) comprising consortium partners: Tennessee State University (TSU), Caltech, Meharry Medical College (MMC), North East Ohio Universities College of Medicine (NEOUCOM), Oak Ridge National Laboratior (ORNL), and the University of Southwest Louisiana (USL). MMC and EOUCOM provided the experimental data. A team of eight (8) researchers along with (5) undergraduate students and twelve (12) graduate students conducted research at the Center. The CNE also supported a Ph.D student at NEOUCOM. CNE conducted research in various aspects of neural computation and applied these techniques to dynamic control aircraft, (helicopter), signal classification (image processing), spatial navigation (mobile robot) medical diagnosis and oscillatory hippocampal network (spatial information processing). The research led to the completion of three Masters Theses and three senior projects; one book chapter, four referenced chapters, six non-referenced conference papers and three new grants in the areas of communication, condition based maintenance and data acquisition system for medical diagnosis.

This Fourth annual Report provides an overview of the research activities in the Center for Neural Engineering (CNE) comprising consortium partners: Tennessee State University (TSU), Caltech, Meharry Medical College (MMC), North East Ohio Universities College of Medicine (NEOUCOM), Oak Ridge National Laboratior (ORNL), and the University of Southwest Louisiana (USL). MMC and EOUCOM provided the experimental data. A team of eight (8) researchers along with (5) undergraduate students and twelve (12) graduate students conducted research at the Center. The CNE also supported a Ph.D student at NEOUCOM. CNE conducted research in various aspects of neural computation and applied these techniques to dynamic control aircraft, (helicopter), signal classification (image processing), spatial navigation (mobile robot) medical diagnosis and oscillatory hippocampal network (spatial information processing). The research led to the completion of three Masters Theses and three senior projects; one book chapter, four referenced chapters, six non-referenced conference papers and three new grants in the areas of communication, condition based maintenance and data acquisition system for medical diagnosis.

Blast barrier walls have been shown to reduce blast loads on structures, especially in urban environments. Analysis of existing test and simulation data for blast barrier response has revealed that a need still exists to determine the bounds of the problem and produce a fast-running accurate model for the effects of barrier walls on blast wave propagation. Since blast experiments are very time intensive and extremely cost prohibitive, it is vital that computational capabilities be developed to generate the required data set that can be utilized to produce simplified design tools. The combination of high fidelity model-based simulation with artificial neural network techniques is proposed in this paper to manage the challenging problem. The proposed approach is demonstrated to estimate the peak pressure, impulse, time of arrival, and time of duration of blast loads on buildings protected by simple barriers, using data generated from validated hydrocode simulations. Once verified and validated, the proposed neural-network model-based simulation procedure would provide a very efficient solution to predicting blast loads on the structures that are protected by blast barrier walls.

The use of continuum models for the analysis of discrete built-up complex aerospace structures is an attractive idea especially at the conceptual and preliminary design stages. But the diversity of available continuum models and hard-to-use qualities of these models have prevented them from finding wide applications. In this regard, Artificial Neural Networks (ANN or NN) may have a great potential as these networks are universal approximators that can realize any continuous mapping, and can provide general mechanisms for building models from data whose input-output relationship can be highly nonlinear. The ultimate aim of the present work is to be able to build high fidelity continuum models for complex aerospace structures using the ANN. As a first step, the concepts and features of ANN are familiarized through the MATLAB NN Toolbox by simulating some representative mapping examples, including some problems in structural engineering. Then some further aspects and lessons learned about the NN training are discussed, including the performances of Feed-Forward and Radial Basis Function NN when dealing with noise-polluted data and the technique of cross-validation. Finally, as an example of using NN in continuum models, a lattice structure with repeating cells is represented by a continuum beam whose properties are provided by neural networks.

Recently neural networks have been applied in antenna modeling where the role of the network is not just for black-box modeling. This paper highlights that aspect of neural networks from the antenna engineering application point of view.

Recently neural networks have been applied in antenna modeling where the role of the network is not just for black-box modeling. This paper highlights that aspect of neural networks from the antenna engineering application point of view.

Recently neural networks have been applied in antenna modeling where the role of the network is not just for black-box modeling. This paper highlights that aspect of neural networks from the antenna engineering application point of view.

Recently neural networks have been applied in antenna modeling where the role of the network is not just for black-box modeling. This paper highlights that aspect of neural networks from the antenna engineering application point of view.

Three graduate students and one undergraduate student conducted research in the Center for Neural Engineering in the areas of biologically motivated neural networks. Their research topics are: (1) developing frequency dependent oscillatory neural networks; (2) long term pontentiation learning rules as applied to spatial navigation; (3) design and build a servo joint robotic arm and (4) neural network based prothesis control. One graduate student published a paper on 'dynamic current-voltage characteristics in neuronal dendrites' in WCNN 95 conferences proceeding.

Four graduate students and one undergraduate student conducted research in the Center for Neural Engineering in the following areas: (1) developing frequency dependent oscillatory neural network architecture for spatial information processing. (2) long term potentiation learning rule as applied to spatial navigation. (3) design and build a servo joint-based robotic arm. (4) design of a neural controller for the control of inverted pendulum and (5) design of intelligent flight control system for helicopter roll-axis. One undergraduate and one graduate student graduate as a result of this award.

Four graduate students and one undergraduate student conducted research in the Center for Neural Engineering in the following areas: (1) developing frequency dependent oscillatory neural network architecture for spatial information processing. (2) long term potentiation learning rule as applied to spatial navigation. (3) design and build a servo joint-based robotic arm. (4) design of a neural controller for the control of inverted pendulum and (5) design of intelligent flight control system for helicopter roll-axis. One undergraduate and one graduate student graduate as a result of this award.

We consider the deterministic evolution of a time-discretized spiking network of neurons with connection weights having delays, modeled as a discretized neural network of the generalized integrate and fire (gIF) type. The purpose is to study a class of algorithmic methods allowing to calculate the proper parameters to reproduce exactly a given spike train generated by an hidden (unknown) neural network. This standard problem is known as NP-hard when delays are to be calculated. We propose here a reformulation, now expressed as a Linear-Programming (LP) problem, thus allowing to provide an efficient resolution. This allows us to "back-engineer" a neural network, i.e. to find out, given a set of initial conditions, which parameters (i.e., connection weights in this case), allow to simulate the network spike dynamics. More precisely we make explicit the fact that the back-engineering of a spike train, is a Linear (L) problem if the membrane potentials are observed and a LP problem if only spike times are observed, with a gIF model. Numerical robustness is discussed. We also explain how it is the use of a generalized IF neuron model instead of a leaky IF model that allows us to derive this algorithm. Furthermore, we point out how the L or LP adjustment mechanism is local to each unit and has the same structure as an "Hebbian" rule. A step further, this paradigm is easily generalizable to the design of input-output spike train transformations. This means that we have a practical method to "program" a spiking network, i.e. find a set of parameters allowing us to exactly reproduce the network output, given an input. Numerical verifications and illustrations are provided.

We consider the deterministic evolution of a time-discretized spiking network of neurons with connection weights having delays, modeled as a discretized neural network of the generalized integrate and fire (gIF) type. The purpose is to study a class of algorithmic methods allowing to calculate the proper parameters to reproduce exactly a given spike train generated by an hidden (unknown) neural network. This standard problem is known as NP-hard when delays are to be calculated. We propose here a reformulation, now expressed as a Linear-Programming (LP) problem, thus allowing to provide an efficient resolution. This allows us to "back-engineer" a neural network, i.e. to find out, given a set of initial conditions, which parameters (i.e., connection weights in this case), allow to simulate the network spike dynamics. More precisely we make explicit the fact that the back-engineering of a spike train, is a Linear (L) problem if the membrane potentials are observed and a LP problem if only spike times are observed, with a gIF model. Numerical robustness is discussed. We also explain how it is the use of a generalized IF neuron model instead of a leaky IF model that allows us to derive this algorithm. Furthermore, we point out how the L or LP adjustment mechanism is local to each unit and has the same structure as an "Hebbian" rule. A step further, this paradigm is easily generalizable to the design of input-output spike train transformations. This means that we have a practical method to "program" a spiking network, i.e. find a set of parameters allowing us to exactly reproduce the network output, given an input. Numerical verifications and illustrations are provided.

We consider the deterministic evolution of a time-discretized spiking network of neurons with connection weights having delays, modeled as a discretized neural network of the generalized integrate and fire (gIF) type. The purpose is to study a class of algorithmic methods allowing to calculate the proper parameters to reproduce exactly a given spike train generated by an hidden (unknown) neural network. This standard problem is known as NP-hard when delays are to be calculated. We propose here a reformulation, now expressed as a Linear-Programming (LP) problem, thus allowing to provide an efficient resolution. This allows us to "back-engineer" a neural network, i.e. to find out, given a set of initial conditions, which parameters (i.e., connection weights in this case), allow to simulate the network spike dynamics. More precisely we make explicit the fact that the back-engineering of a spike train, is a Linear (L) problem if the membrane potentials are observed and a LP problem if only spike times are observed, with a gIF model. Numerical robustness is discussed. We also explain how it is the use of a generalized IF neuron model instead of a leaky IF model that allows us to derive this algorithm. Furthermore, we point out how the L or LP adjustment mechanism is local to each unit and has the same structure as an "Hebbian" rule. A step further, this paradigm is easily generalizable to the design of input-output spike train transformations. This means that we have a practical method to "program" a spiking network, i.e. find a set of parameters allowing us to exactly reproduce the network output, given an input. Numerical verifications and illustrations are provided.

We consider the deterministic evolution of a time-discretized spiking network of neurons with connection weights having delays, modeled as a discretized neural network of the generalized integrate and fire (gIF) type. The purpose is to study a class of algorithmic methods allowing to calculate the proper parameters to reproduce exactly a given spike train generated by an hidden (unknown) neural network. This standard problem is known as NP-hard when delays are to be calculated. We propose here a reformulation, now expressed as a Linear-Programming (LP) problem, thus allowing to provide an efficient resolution. This allows us to "back-engineer" a neural network, i.e. to find out, given a set of initial conditions, which parameters (i.e., connection weights in this case), allow to simulate the network spike dynamics. More precisely we make explicit the fact that the back-engineering of a spike train, is a Linear (L) problem if the membrane potentials are observed and a LP problem if only spike times are observed, with a gIF model. Numerical robustness is discussed. We also explain how it is the use of a generalized IF neuron model instead of a leaky IF model that allows us to derive this algorithm. Furthermore, we point out how the L or LP adjustment mechanism is local to each unit and has the same structure as an "Hebbian" rule. A step further, this paradigm is easily generalizable to the design of input-output spike train transformations. This means that we have a practical method to "program" a spiking network, i.e. find a set of parameters allowing us to exactly reproduce the network output, given an input. Numerical verifications and illustrations are provided.

An Artificial neural network ANN is an information processing hypothesis that is stimulated by the way natural nervous system, such as brain, process information. The using of artificial neural network in civil engineering is getting more and more credit all over the world in last decades. This soft computing method has been shown to be very effective in the analysis and solution of civil engineering problems. It is defined as a body which works out the more and more complex problem through sequential algorithms. It is designed on the basis of artificial intelligence which is proficient of storing more and more information's. In this work, we have investigated the various architectures of ANN and their learning process. The artificial neural network based method was widely applied to the civil engineering because of the strong non linear relationship between known and un known of the problems. They come with good modelling in areas where conventional approaches finite elements, finite differences etc. require large computing resources or time to solve problems. These includes to study the behaviour of building materials, structural identification and control problems, in geo technical engineering like earthquake induced liquefaction potential, in heat transfer problems in civil engineering to improve air quality, in transportation engineering like identification of traffic problems to improve its flexibility , in construction technology and management to estimate the cost of buildings and in building services issues like analyzing the water distribution network etc. Researches reveals that the method is realistic and it will be fascinated for more civil engineering applications.  by Vikash Singh | Samreen Bano | Anand Kumar Yadav | Dr. Sabih Ahmad "Feasibility of Artificial Neural Network in Civil Engineering"  Published in International Journal of Trend in Scientific Research and Development (ijtsrd), ISSN: 2456-6470, Volume-3 | Issue-3 , April 2019,  URL: https://www.ijtsrd.com/papers/ijtsrd22985.pdf Paper URL: https://www.ijtsrd.com/engineering/civil-engineering/22985/feasibility-of-artificial-neural-network-in-civil-engineering/vikash-singh

The main objectives of the Center for Neural Engineering are: (1) To advance the understanding of biologically-motivated neural network systems through inter-disciplinary basic research; (2) To develop the highest quality undergraduate and graduate curricula in neural computing and engineering that will serve as a role model for other institutions; (3) To provide pre-graduate and post-graduate training for students in a nationally and internationally recognized basic and applied research and development environment focusing on critical present and future technologies; and (4) To broaden educational and career development opportunities for minorities and women.

The main objectives of the Center for Neural Engineering are: (1) To advance the understanding of biologically-motivated neural network systems through inter-disciplinary basic research; (2) To develop the highest quality undergraduate and graduate curricula in neural computing and engineering that will serve as a role model for other institutions; (3) To provide pre-graduate and post-graduate training for students in a nationally and internationally recognized basic and applied research and development environment focusing on critical present and future technologies; and (4) To broaden educational and career development opportunities for minorities and women.

This work describes a methodology to combine logic-based systems and connectionist systems. Our approach uses finite truth valued {\L}ukasiewicz logic, where we take advantage of fact what in this type of logics every connective can be define by a neuron in an artificial network having by activation function the identity truncated to zero and one. This allowed the injection of first-order formulas in a network architecture, and also simplified symbolic rule extraction. Our method trains a neural network using Levenderg-Marquardt algorithm, where we restrict the knowledge dissemination in the network structure. We show how this reduces neural networks plasticity without damage drastically the learning performance. Making the descriptive power of produced neural networks similar to the descriptive power of {\L}ukasiewicz logic language, simplifying the translation between symbolic and connectionist structures. This method is used in the reverse engineering problem of finding the formula used on generation of a truth table for a multi-valued {\L}ukasiewicz logic. For real data sets the method is particularly useful for attribute selection, on binary classification problems defined using nominal attribute. After attribute selection and possible data set completion in the resulting connectionist model: neurons are directly representable using a disjunctive or conjunctive formulas, in the {\L}ukasiewicz logic, or neurons are interpretations which can be approximated by symbolic rules. This fact is exemplified, extracting symbolic knowledge from connectionist models generated for the data set Mushroom from UCI Machine Learning Repository.

This article is from Frontiers in Neuroscience , volume 8 . Abstract Efficient Analog-Digital Converters (ADC) are one of the mainstays of mixed-signal integrated circuit design. Besides the conventional ADCs used in mainstream ICs, there have been various attempts in the past to utilize neuromorphic networks to accomplish an efficient crossing between analog and digital domains, i.e., to build neurally inspired ADCs. Generally, these have suffered from the same problems as conventional ADCs, that is they require high-precision, handcrafted analog circuits and are thus not technology portable. In this paper, we present an ADC based on the Neural Engineering Framework (NEF). It carries out a large fraction of the overall ADC process in the digital domain, i.e., it is easily portable across technologies. The analog-digital conversion takes full advantage of the high degree of parallelism inherent in neuromorphic networks, making for a very scalable ADC. In addition, it has a number of features not commonly found in conventional ADCs, such as a runtime reconfigurability of the ADC sampling rate, resolution and transfer characteristic.